OpenAI discontinued Sora: the app in April 2026 and the API in September. HeyOz runs the models people moved to — Seedance 2.5, Kling 4, Veo 3.1, MiniMax H3 and more — in one studio built for making ads rather than filling a feed.
Product films, cinematic spots, animation and fashion, all made on HeyOz since Sora closed. Same briefs, different engines.
Your prompts still work. Only the engine changes.
Open the video generator and choose the engine that fits the shot. The picker shows each model's resolution, length and whether it generates audio, so the choice takes seconds.
Concrete, structured prompts transfer directly: subject, action, environment, camera, treatment. Paste what worked in Sora and add a line about the sound you want.
Set your length, ratio and resolution, then generate. Most models here return the clip with audio already in it, sized for the channel you are posting to.
Sora was one model inside a consumer app. This is a shelf of production models inside an ad studio.
OpenAI retired the Sora app and website on 26 April 2026 and switched off the API on 24 September 2026. There is no paid tier that brings it back. If you still have work inside it, export it from the Sunset page before the window closes.
Seedance 2.5 runs four to thirty seconds at up to 1080p and decides sound and picture together — dialogue with lip sync, foley and room tone, cut to the frame. No separate audio step before a clip can be posted.
Veo 3.1 for cinematic shots, Kling 4 for long takes, MiniMax H3 for 2K and on-screen type, Runway Gen-4.5 for physics, Wan 3.0 for thirty-second runs. Switch model without switching tool, because no single engine wins every brief.
Sora was shut down as a resource-heavy consumer app with little revenue. HeyOz is a paid product for people making ads, which is a different kind of thing to build a workflow on.
And the parts a consumer video app was never going to cover.
Steam, pour and sizzle, where the physics has to read as real. The shots that sell an appetite rather than describe a product.
Your actual item, with its real material and finish, lit and moved the way a studio would do it. Reference your photography and it comes back as yours.
Illustrated and 3D characters that hold their design across a campaign, so a mascot stays on model from the first asset to the last.
Type, texture and shape in motion, for the brand moments where there is no product to film and the idea has to carry the frame.
No. OpenAI discontinued the Sora web and app experiences on 26 April 2026, and the Sora API on 24 September 2026. The shutdown was announced in March 2026 as part of a shift toward business tools.
OpenAI's help centre points users to the Sunset page to export their content, and says data will be permanently deleted after any final export window closes. Export anything you want to keep as soon as you can.
OpenAI states that purchased ChatGPT or Sora credits can still be used for Codex. For refunds, their help centre points to the standard ChatGPT subscription refund process.
It depends which part you relied on. For long takes with sound, Seedance 2.5 runs up to thirty seconds with audio and lip sync generated alongside the picture. For cinematic look, Veo 3.1. For 2K output and legible on-screen type, MiniMax H3.
Most of them. Seedance 2.5, Veo 3.1, Kling and Wan 3.0 generate sound with the picture. Runway Gen-4.5 is the exception — it is picture only, so plan on adding audio afterwards.
Yes. Concrete, structured prompts transfer well because these models reward the same specificity: subject, action, environment, camera behaviour and treatment. The main addition is describing the sound, since most engines here generate it.
It varies by engine. Seedance 2.5 and Wan 3.0 reach thirty seconds, MiniMax H3 runs to sixteen, Veo 3.1 to eight. The model picker shows the range before you generate.
Yes, and that is the main difference. Sora was a consumer app for making and sharing clips. HeyOz starts from your brand, makes the formats ad channels want, and sizes them for where you are posting.
Pick the right engine for the job — all in one place.
Anyone can hand you a video model. HeyOz turns it into your ad.
HeyOz reads your brand and keeps every asset on-color, on-voice, on-message.
Seedance, Veo, Kling, GPT Image, and more — no juggling subscriptions.
Sized and formatted for Meta and TikTok, straight out of the box.
Not just a model — templates, avatars, and an agent that does it all.
Every model that replaced it is already in the studio. Bring your prompts and pick up where you left off.
